Tuesday, March 27, 2007
Lab 6
http://mason.gmu.edu/~adement/Lab6_v2.html
(1) OK. I had issues. I used the 20% alpha to generate the color gradients; however, even after 5 states were selected, the color did not hit "100%" and remain constant. Apparently, the way the transparency is generated by the computer, the addition of more transparent layers just continues to deepen the color. As a result, up to 8 shades are possible if all maps are selected. Beyond 5 years, it becomes nearly impossible to distinguish between all the colors.
So...I went ahead and did all 8 maps but I only put 5 colors in the legend and I told the user to only pick 5. While the user could still check all 8 boxes (proving their inability to follow directions), the map becomes increasingly useless and the disobedient reader must extrapolate the colors (which serves them right). I couldn't figure out how to make the software police their actions and limit them to only picking 5 of the 8 years.
(2) Because of this same color problem, creating the legend was problematic. For example, 4 overlapped 20% alpha symbols on the map yields a significantly different color than one 80% alpha graphics object in the legend. As a result, to replicate the map shades in the legend, I had to generate symbols and overlap them just like on the map rather than just drawing objects and using the standard fill tool with different alphas (40%, 60%, etc.)
(3) I did the mapping with a plain white background and then decided to add a color to the background afterward. Turns out that because of the transparency of the symbols, adding a color behind them completely changes the visual perception of color for the symbols. This can have some shocking results. Fortunately, I found that a sick yellow shade behind the maps actually provided better gradient differentiation than using white alone. Picking the background color before the symbol color, however, is a design concern that would be better addressed at the start of a project.
(4) I wouldn't say this is my worst map but because of the color issues, it's now in the top 3 along with dot density and the Maryland cows. I like the whole faux-GIS idea, though.
(1) OK. I had issues. I used the 20% alpha to generate the color gradients; however, even after 5 states were selected, the color did not hit "100%" and remain constant. Apparently, the way the transparency is generated by the computer, the addition of more transparent layers just continues to deepen the color. As a result, up to 8 shades are possible if all maps are selected. Beyond 5 years, it becomes nearly impossible to distinguish between all the colors.
So...I went ahead and did all 8 maps but I only put 5 colors in the legend and I told the user to only pick 5. While the user could still check all 8 boxes (proving their inability to follow directions), the map becomes increasingly useless and the disobedient reader must extrapolate the colors (which serves them right). I couldn't figure out how to make the software police their actions and limit them to only picking 5 of the 8 years.
(2) Because of this same color problem, creating the legend was problematic. For example, 4 overlapped 20% alpha symbols on the map yields a significantly different color than one 80% alpha graphics object in the legend. As a result, to replicate the map shades in the legend, I had to generate symbols and overlap them just like on the map rather than just drawing objects and using the standard fill tool with different alphas (40%, 60%, etc.)
(3) I did the mapping with a plain white background and then decided to add a color to the background afterward. Turns out that because of the transparency of the symbols, adding a color behind them completely changes the visual perception of color for the symbols. This can have some shocking results. Fortunately, I found that a sick yellow shade behind the maps actually provided better gradient differentiation than using white alone. Picking the background color before the symbol color, however, is a design concern that would be better addressed at the start of a project.
(4) I wouldn't say this is my worst map but because of the color issues, it's now in the top 3 along with dot density and the Maryland cows. I like the whole faux-GIS idea, though.
# posted by Anne Dement @ 12:23 PM 
